Forecasting - the Cinderella of corporate management?

Forecasting processes are costly. And because changes in the business environment are occurring more frequently and more unexpectedly, clients are increasingly questioning their internal forecasting processes. Agile planning requires lean and fast processes - and digital helpers.

What are the benefits of analytical and predictive models?

Mostly, planning processes are very elaborate and can also be politically motivated. In practice, the forecasting process is often only dealt with in passing. But because external influences today affect corporate structures more quickly, more frequently and more unexpectedly, planning must be updated more often. This requires lean and fast processes - i.e. predictive models. s-peers has developed such models for accounting and controlling. Predictive support is ideally suited for the financial sector because standard systems and processes exist here that deliver very good data quality. Based on past data, predictive systems recognise data patterns and derive predictions from them - with very high accuracy. The system causes predictions to be made faster, more efficiently and thus more frequently. This enables management to work more proactively and to intervene quickly and specifically when necessary.

s-peers models have a prediction accuracy of 95 to 99 per cent

s-peers uses state-of-the-art algorithms to create models. An analysis and prediction model is created based on so-called training data - e.g. turnover from 2011-2017. The model is then checked with validation data - e.g. turnover from 2018. The best performing algorithm is then tested with independent evaluation data.

Methodology combined with gut feeling

Automation objectifies the forecasting processes. However, IT-supported forecasts can be wonderfully combined with individually created plans
. Human experience and impulses, combined with calculated models, provide a realistic picture and thus a good
basis for decisions. There is a great desire for predictive projects. But automated predictions are only as good as the underlying information. Therefore, you should start with an area that already has a very good database. We recommend starting with the area of accounting in order to establish the first added values. Here, there are tried-and-tested standard systems, most of which are based on a uniform foundation.

The advantages of automated forecasting processes for the company

• More capacity remains for value-added activities.
• The results become more precise, the transparency greater.
• Interrelationships and cross-functional dependencies become clear.
• The ability to react increases significantly.
• Forecasts are objectified and reflect a fact-based picture.
• The performance of the algorithms can be tested by simulation.
• The algorithms of s-peers are reusable and can be used several times.
• There is a wide range of evaluation options in assured quality.

The use of automated forecasting processes for the company

The predictive approach improves and accelerates decisions based on quantitative, more differentiated findings: Through the holistic view of cause-effect chains, interrelationships and cross-functional dependencies of a wide range of business areas can also be taken into account much better.

• Other areas of application for automated forecasting models are, for example:
• Improving the stock of goods in the retail trade
• More precise scheduling in production
• Personnel planning
• Optimisation of delivery times
• Price and conditions policy
• Assortment decisions

Conclusion

The use of scientific, statistical models in forecasting is recognized and proven - and corporate management with the help of predictive controlling and accounting holds enormous potential. Success factors for successful implementation are the quality or value
of the data, the correct use of the algorithms, and their constant validation. Get the Cinderella out from behind the stove and test what it is capable of.